Make it clear to a user that we're skipping collections after blank XRC

[shihab-dls]

Fixes issue #784
Added the hotfix @DominicOram wrote in the beamline deployment, and a unit test to test the ispyb callback stop event.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 86.94%. Comparing base (2ff4607) to head (a91d86f).
Report is 1 commits behind head on main.

Additional details and impacted files

@@           Coverage Diff           @@
##             main     #814   +/-   ##
=======================================
  Coverage   86.94%   86.94%           
=======================================
  Files         102      102           
  Lines        6968     6972    +4     
=======================================
+ Hits         6058     6062    +4     
  Misses        910      910           

Components	Coverage Δ
i24 SSX	72.84% <ø> (ø)
hyperion	96.33% <ø> (ø)
other	96.62% <100.00%> (+<0.01%)	⬆️

[shihab-dls]

I assigned the conftest runstop event test_gridscan3d_stop_document_with_crystal_exception the values:
"time": 1666604299.6149616
"uid": "65b2bde5-5740-42d7-9047-e860e06fbe15"
arbitrarily, as they are required; I think this is fine, but commenting just in case it is not.

[DominicOram]

Thanks for the test. As discussed, I think there's a cleaner solution

[DominicOram]

Should: As discussed, checking strings is bad, we can use type_and_message_from_reason here I think

[DominicOram]

I assigned the conftest runstop event test_gridscan3d_stop_document_with_crystal_exception the values: "time": 1666604299.6149616 "uid": "65b2bde5-5740-42d7-9047-e860e06fbe15" arbitrarily, as they are required; I think this is fine, but commenting just in case it is not.

Yep, this is fine, thanks!

[DominicOram]

Thank you! Much cleaner than my hotfix. Optional change, take it or leave it.